Tài liệu Biodiversity Response to Climate Change in the Middle Pleistocene ppt

Biodiversity Response to Climate Changein the Middle Pleistocene The Porcupine Cave Fauna from Colorado... London, England © 2004 by The Regents of the University of California Library o

Biodiversity Response to Climate Change

in the Middle Pleistocene

The publisher gratefully acknowledges the generous contribution to this book provided by the General Endowment Fund of the University of California Press Associates.

Biodiversity Response to Climate Change

in the Middle Pleistocene

The Porcupine Cave Fauna from Colorado

Berkeley and Los Angeles, California

University of California Press, Ltd.

London, England

© 2004 by The Regents of the University of California

Library of Congress Cataloging-in-Publication Data

Biodiversity response to climate change in the middle Pleistocene : the Porcupine Cave fauna from Colorado / edited by Anthony D Barnosky

p cm.

Includes bibliographical references and index.

ISBN 0-520-24082-0 (cloth : alk paper)

1 Vertebrates, Fossil—Colorado—Park County 2 Paleontology— Pleistocene 3 Paleoecology—Colorado—Park County 4 Paleo- ecology—Pleistocene 5 Climatic changes—Environmental aspects— Colorado—Park County—History 6 Animals, Fossil—Colorado— Park County I Title: Porcupine Cave fauna from Colorado.

II Barnosky, Anthony D., 1952–

To Don Rasmussen, who with his son Larryfound the fossil deposits in Porcupine Cave,and whose unfailing enthusiasm fordiscovery, excavation, and workingwith other scientists and volunteerswas essential in moving the projectfrom concept to reality.

PREFACE ix

ACKNOWLEDGMENTS xi

LIST OF CHAPTER APPENDIXES xiii

LIST OF FIGURES xv

LIST OF TABLES xix

ABBREVIATIONS AND DEFINITIONS xxi

PA R T O N E

The Discovery and Distribution

of Fossils

1 Climate Change, Biodiversity,

and Ecosystem Health: The Past

as a Key to the Future 3

Anthony D Barnosky

2 The Pleistocene Fossils of

Porcupine Cave, Colorado:

Spatial Distribution and

Taphonomic Overview 6

Anthony D Barnosky,

Christopher J Bell, Robert G

Raynolds, and Louis H Taylor

3 The Modern Environment,

Flora, and Vegetation of

South Park, Colorado 27

David J Cooper

4 The Historical Context of

Porcupine Cave: American

Indians, Spaniards, Government

Surveyors, Prospectors, Ranchers,

Cavers, and Paleontologists in

South Park, Colorado 39

Geraldine J Rasmussen, Kirk

Branson, and John O McKelvy

5 The Geology and Speleogenesis

of Porcupine Cave 51Robert G Raynolds

6 Magnetostratigraphic Constraints

on the Age of PleistoceneFossiliferous Strata in PorcupineCave’s DMNH Velvet RoomExcavation 57

S Julio Friedmann and Robert G Raynolds

7 Age and Correlation of Key FossilSites in Porcupine Cave 64Anthony D Barnosky andChristopher J Bell

8 Biology of Wood Rats as CaveDwellers and Collectors 74Robert B Finley Jr

9 Paleopathology and TaphonomicModification of Mammalian Bonesfrom Porcupine Cave 82

C Suzane Ware and Elaine Anderson

PA R T T WOSystematic Accounts of Taxa

10 A Summary of Fossilized Species

in Porcupine Cave 95Anthony D Barnosky

11 Synopsis of the Herpetofaunafrom Porcupine Cave 117Christopher J Bell, Jason J Head,and Jim I Mead

12 The Early and Middle Pleistocene Avifauna fromPorcupine Cave 127Steven D Emslie

13 The Carnivora from PorcupineCave 141

Elaine Anderson

14 Middle Pleistocene (Irvingtonian)

Ochotona (Lagomorpha:

Ochotonidae) from Porcupine Cave 155Jim I Mead, Margarita Erbajeva,and Sandra L Swift

15 Leporidae of the DMNH Velvet Room Excavations and Mark’s Sink 164Colleen N Baxter

16 Identification of MiscellaneousMammals from the Pit Locality:Including Soricidae, Leporidae,Geomyoidea 169

Anthony D Barnosky andSamantha S B Hopkins

17 Systematics and FaunalDynamics of Fossil Squirrels from Porcupine Cave 172

H Thomas Goodwin

18 Fossil Wood Rats of Porcupine Cave: Tectonic

or Climatic Controls? 193Charles A Repenning

C O N T E N TS

19 Arvicoline Rodents from

Porcupine Cave: Identification,

Spatial Distribution, Taxonomic

Assemblages, and

Biochronologic Significance 207

Christopher J Bell,

Charles A Repenning,

and Anthony D Barnosky

20 Pliocene and Pleistocene Horses

from Porcupine Cave 264

Effect of Environmental Change

on the Porcupine Cave Fauna

22 Irvingtonian Mammals from the

Badger Room in Porcupine Cave:

Age, Taphonomy, Climate, and

Ecology 295

Alan B Shabel,

Anthony D Barnosky,

Tonya Van Leuvan, Faysal Bibi,

and Matthew H Kaplan

23 Faunal Dynamics of Small

Mammals through the Pit

Sequence 318

Anthony D Barnosky

24 Stable Carbon and OxygenIsotope Analysis of MarmotCheek Teeth from the Pit Locality 327

Robert S Feranec

25 Assessing the Effect of MiddlePleistocene Climate Change

on Marmota Populations from

the Pit Locality 332Anthony D Barnosky, Matthew

H Kaplan, and Marc A Carrasco

26 Effect of Climate Change onTerrestrial Vertebrate Biodiversity

LITERATURE CITED 347 LIST OF CONTRIBUTORS 371 INDEX 373

Since fossil vertebrates were first discovered at Porcupine Cave

on the rim of South Park, Colorado, in 1981, the site has

be-come the world’s most important source of information about

animals that lived in the high elevations of North America in

the middle part of the ice ages, between approximately one

million and 600,000 years ago Beginning in 1985, teams of

scientists and volunteers from three major research institutions

—the Carnegie Museum of Natural History, the Denver

Mu-seum of Nature and Science, and the University of California

Museum of Paleontology—spent some 15 field seasons

exca-vating and studying tens of thousands of fossil specimens that

have opened a window onto past evolutionary and ecological

adjustments This window into the past allows us to visualize

how ongoing global change could affect our living

commu-nities This book reports the results of nearly two decades of

research and has been written to appeal to three overlapping

audiences

The first target audience is made up of scholars, students, and

others interested in paleontology and in how paleontological

data are applied to solving ecological and evolutionary

ques-tions The second audience consists of ecologists and

conser-vationists concerned with understanding and preserving

biodiversity and other natural ecological dynamics To serve

these first two audiences, the book strives to illustrate the

critical role paleontology plays in understanding ecosystem

dynamics, such as the maintenance of biodiversity, and to

document carefully the scientific data from Porcupine Cave

so that this unique data set can be used now and in the future

to illuminate ecological processes

The third audience is the caving community, which has

increasingly used Porcupine Cave and others like it for both

scientific and recreational purposes over the past 20 years.The book endeavors to highlight the importance of the frag-ile but irreplaceable paleontological resources to be found incaves

The book is divided into three parts The chapters in part 1articulate some scientific questions that the data from thecave can help answer; document the location, modern envi-ronment, and geological setting of the site as a context inwhich to interpret the fossil data; and relate the history of thediscovery of Porcupine Cave, the spatial distribution and age

of the fossil deposits, and the cause of the accumulation of somany fossils Part 2 documents the identification and occur-rence of various taxonomic groups from the many differentlocalities within the cave Part 3 synthesizes the informationpresented in the other two parts into a series of analyses de-signed to explore the implications of the Porcupine Cave faunafor understanding how terrestrial mountain ecosystems react

in the face of environmental change, how climate changeaffects patterns of biodiversity in mammals, and, in light ofthese processes, how we might expect ecosystems to respond

to human-induced global warming

Given the astounding numbers of fossils that PorcupineCave has produced—more than 20,000 specimens have beenidentified, and many times that number are stored in mu-seum drawers awaiting identification—it is impossible to ex-plore all their implications in a single publication This bookshould be viewed as a foundation for further research ratherthan the final word on the matter We hope that the data andideas presented herein stimulate debate and provide impetusfor a new cohort of scholars to continue the work we have justbegun

P R E FAC E

Excavating, analyzing, and publishing the Porcupine Cave

data has been an arduous task that has taken nearly 20 years

and involved more than 30 scientists, more than 100 field

hands, and the cooperation of the three major museums where

specimens reside: the Carnegie Museum of Natural History

(CM), the Denver Museum of Nature and Science (DMNH),

and the University of California Museum of Paleontology

(UCMP) Thanks are due to all who lent a hand, and especially

to the following individuals and institutions

The kindness of Frank and Connie McMurry (McMurry

Land and Livestock Company) in allowing us to excavate in

their cave and spend field seasons at their cow camp made the

whole project possible I am deeply indebted to them The

project would also not have been possible without financial

support from the U.S National Science Foundation (grants

BSR-9196082 in the early years and EAR-9909353 during the

synthesis stage), the UCMP, the CM, and the DMNH

Don Rasmussen spearheaded the excavation teams for

many years and contributed in innumerable ways to the

proj-ect He has been a delight to work with Two colleagues who

contributed essential data to this project died before they saw

the fruits of their labors: Vic Schmidt (paleomagnetics) and

Elaine Anderson (carnivores) Memories of days in the cave

and nights at the campfire with them live on Betty Hill of the

CM, Logan Ivy of the DMNH, and Pat Holroyd of the UCMP

were extremely helpful in arranging loans of specimens and

dealing with sometimes overwhelming curatorial matters

Paul Koch graciously ran isotope samples in his lab at the

Uni-versity of California, Santa Cruz Karen Klitz of the UniUni-versity

of California Museum of Vertebrate Zoology (MVZ) prepared

some of the illustrations, and the MVZ was an essential

re-source for specimen identification

It is impossible to name here the more than 100 volunteers,

students, and employees who helped excavate the deposits and

pick matrix, but I am grateful to them all Many of them were

members of the Colorado Grotto of the National

Speleologi-cal Society or the Western Interior PaleontologiSpeleologi-cal Society

Hazel Barton led the cartographic efforts to produce detailedmaps of the cave, and I thank her for making her beautifulmap available

As editor, my job was made easier by the contributors to thisbook, many of whom waited patiently after submission oftheir manuscripts for the whole package to come together.Special thanks are due to Chris Bell, who has been with theproject since the early 1990s and who picked most of the ma-trix from the Pit Several scientists gave of their time in pro-viding detailed reviews of various chapters: Elaine Anderson,Jill Baron, Chris Bell, Annalisa Berta, Doug Burbank, JimBurns, Emmet Evanoff, Bob Feranec, Tom Goodwin, FredGrady, Elizabeth Hadly, Bill Harbert, Art Harris, R Lee Lyman,Bruce MacFadden, Bob Martin, David Polly, Don Rasmussen,Bruce Rothschild, Dennis Ruez Jr., Eric Scott, Alan Shabel,David Steadman, Tom Stidham, Tom Van Devender, BlaireVan Valkenburgh, Alisa Winkler, Bill Wyckoff, and RichardZakrzewski To them, and to several reviewers who wished toremain anonymous, I give thanks Gratitude is also extended

to Timothy Heaton and Karel Rogers, who read the entiremanuscript of the book and provided useful comments.Discussions with my graduate and postdoctoral studentshave been intellectually stimulating Alan Shabel’s deep think-ing about ecology and willingness to lend a hand as neededwere a great benefit, as were discussions with Marc Carrasco,Edward Davis, Bob Feranec, Samantha Hopkins, and BrianKraatz Faysal Bibi curated many of the UCMP Pit specimensand helped produce some of the spreadsheets used in theanalyses

Above all, I thank my wife, Liz Hadly, for her help bothscientifically and with living life, and my children, Emmaand Clara, who make me think about why biodiversity might

be important for future generations

A D Barnosky Palo Alto, California

AC K N OW L E D G M E N TS

9.1 Specimens from Porcupine Cave showing taphonomic

11.1 Amphibian and reptile specimens recovered from

17.1 List of specimens examined for the four most

19.1 Specimens of arvicoline rodents from

22.1 Repository numbers of all specimens used in

C H A P T E R A P P E N D I X E S

1.1 Per-hundred-year temperature change values for

global warming events plotted against the interval

of time over which the temperature change was

from middle Pleistocene strata in the CM Velvet

than 0.25, more than 1.25, and more than 2.54 cm

of precipitation at the Antero Reservoir weather

flowering paintbrush and other herbaceous

Kobresia simpliciuscula 36

Ptilagrostis porteriin North America and the closely

geographic features, landmarks, Bautista de Anza’s

F I G U R E S

5.3 Stratigraphic section at Porcupine Cave 53

ochotonids, leporids, sciurids, and geomyids at

stratigraphic sequences with global climate changes

Canis latransskull 84

Canis latransfemur 85

9.4 Lepussp (hare or jackrabbit), DMNH 42146,

9.5 Lepussp (hare or jackrabbit), DMNH 42148,

9.6 Lepussp (hare or jackrabbit), DMNH 20052, right

9.7 Canis latrans(coyote), DMNH 30076, skull from the

(ground squirrel), DMNH 41425, right innominate

9.9 Canis latrans(coyote), DMNH 26646, right femur

N arquata(Far Eastern or Eurasian Curlew) 130

maxilla with P4 from Mark’s Sink, and DMNH 34570,

containing Ochotona of Hemphillian, Blancan, and

Cynomys 174

scatterplot of trigonid width versus length

width of p4 of S cf S elegans and relative frequency

17.4 Specimens of ?Cynomys andersoni sp nov. 180

deflection of hypoconid on m3 among four samples

of prairie dogs; discrimination of modern C.

gunnisoni and C leucurus based on discriminant

premolars (right p2, DMNH 27524; right p3 or p4,

from Porcupine Cave, compared with like elements

20.10 Bivariate plot of measurements of astragali from

Porcupine Cave, compared with like elements of

20.11 First phalanges of Equus (Hemionus) sp from

20.12 First phalanx of Equus (Hemionus) sp from

Porcupine Cave compared with first phalanx of

E conversidens 275

20.14 Right m1 or m2 (DMNH 41172) from Mark’s Sink in

20.15 Bivariate plot of measurements of second phalanges

of small Equus from Porcupine Cave, compared

with like elements of hemionines from other

Ovis,and Antilocapridae 286

temperature (winter) maxima on the climate space

shared by the indicator taxa, using Spermophilus elegans and S richardsonii 309

plotted against NISP for the Porcupine Cave Pit

Porcupine Cave Pit taxa broken out by

through time in the Porcupine Cave Pit sequence,

through time in the Porcupine Cave Pit sequence,

molars with four triangles versus those with five or

expected from the Coleman rarefaction analysisagainst the observed richness for stratigraphic

25.2 Log of the area (AP × T) of the upper P4 plotted as a

2.1 Names and locality numbers of major vertebrate

applicable only within grids to correlative horizons

labeled by letters traceable across grids in DMNH

South Park, compared with data from Fairbanks,

Alaska, the Colorado Front Range, and the

Room, Cramped Quarters, Crystal Room, and

Gypsum Room SE Corner, KU Digs 1 and 3, and

Entrance, Velvet Room Last Chance Pit, Velvet Room

10.12 Faunal list for upper six horizons of the DMNH

10.13 Faunal list for lower six horizons andundifferentiated material from the DMNH

urophasianuscompared to the fossil specimen

compared to the fossil specimen from Porcupine

TA B L E S

13.3 Most common species of carnivores in five areas of

Hemphillian and Blancan) and early Pleistocene

form of pika here termed Ochotona sp near Trout

horizons of the main dig site in the Velvet Room 168

primitive prairie dogs and prairie dog–like ground

of Neotoma represented in the Porcupine Cave

percentages of NISP by level for Neotoma fossils

modern and fossil Antilocapra and fossils from

specimen CM 75510; Oreamnos harringtoni

from Texas; Oreamnos harringtoni from the Grand Canyon, Arizona; Oreamnos americanus;

elements in the Carnegie Museum Badger Room

femora and humeri in the Carnegie Museum

temperatures within the geographic ranges of

maximum monthly precipitation levels based on

mammals compared with the recent historical

community broken down by size class and

22.10 Species deletion and addition information broken

1.77 Ma ago, following the placement of the

boundary in the section at Vrica, Italy, by

International Geological Correlation Project

41 and International Union of Quaternary

Research Subcommission 1d at the 27th

International Geological Congress in

Moscow in 1984 (Bell et al., in press), and

subsequent correlation of the boundary with

the magnetostratigraphic and radiometric

time scales (Cande and Kent, 1995; Berggren

et al., 1995) The use of the term “middle

Pleistocene” in this book is informal and

refers to the middle third of the Pleistocene,

that is, the interval of time from about 600

Ka to 1.2 Ma ago This is not to be confused

with terminology such as “Middle

Pleistocene subseries” (note the uppercase

working definition, adopted by the

International Commission on Stratigraphy,

that encompasses the time interval from

circa 126 to 780 Ka ago (http://micropress

.org/stratigraphy/gssp.htm)

the first upper molar and p4 stands for thefourth lower premolar

Institution abbreviations

Discoveries

TMM Texas Memorial Museum

Paleontology

Measurement abbreviations

portion of tooth

surface

tooth measured perpendicular to AP

PA R T O N E

T H E D I S C OV E RY A N D

D I ST R I B U T I O N O F F O S S I LS

Earth’s climate is getting warmer, and it will probably

tinue to do so over the coming century The emerging

con-sensus is that human activities are stimulating an increase in

global mean temperature that will amount to 1.4–5.8°C by the

year 2100 (Houghton et al., 2001), with 90% probability that

the change will amount to 1.7–4.9°C in the absence of climate

mitigation policies (Wigley and Raper, 2001) Regionally, the

changes will be even greater Average warming for the United

States is predicted to be at least 3°C and possibly as much as

6°C (National Assessment Synthesis Team, 2001) The effects

of some of these changes are already apparent For example,

a warming of approximately 4°C in Alaska since the 1970s has

led to vast expanses of spruce forests being killed by beetles

that reproduce faster in warmer temperatures Roads are

buck-ling and houses are sinking, as what used to be permafrost

thaws seasonally

A growing number of scientists have recognized that global

warming can be expected to affect the few remaining intact,

naturally operating ecosystems on Earth in unpredictable

ways This issue came to widespread attention just over a

decade ago, with the publication of a compendium of papers,

edited by Peters and Lovejoy (1992), concerning the effects

of global warming on biodiversity The effects of climate

change on biodiversity are a matter of concern because

bio-diversity is often associated with ecosystem health

Signifi-cant losses in biodiversity may be analogous to the death of

the canary in the coal mine, which signals that the mine is

no longer safe for humans Though debate continues about

whether “more is better” in terms of numbers of species in

eco-systems (Norton, 1987; Grime, 1997; Tilman, 1997; McCann,

2000), available information suggests that larger numbers of

species help buffer ecosystems in the face of changing

envi-ronments (Loreau et al., 2001) Thus of key concern is the

question of whether climatic warming will reduce

biodiver-sity to the extent that a given ecosystem loses its ability to

maintain the baseline functions that define it Maintaining

these baseline functions is, in fact, integral to an operationaldefinition of ecosystem health In the words of Haskell et al.(1992:9), “An ecological system is healthy if it is stable andsustainable—that is, if it is active and maintains its organi-zation and autonomy over time and is resilient to stress.” Putanother way, the basic question is: at what point do disrup-tions to baseline diversity cause ecosystems to cross functionalthresholds and catastrophically shift their dynamics (Sheffer

et al., 2001)?

Adding to concerns about the effects of climate change

on biodiversity is the fragmentation of previously widespreadbiota by human activities, which itself—probably more sothan climate change—often leads to reduction in species rich-ness As Soulé (1992:xiii) put it, it is simply the wrong time forclimate change “Even if species are able to move quicklyenough to track their preferred climate, they will have to do

so within a major obstacle course set by society’s conversion

of the landscape A species may be impelled to move, butLos Angeles will be in the way” (Peters and Lovejoy, 1992:xviii)

Over the past decade, researchers have continued to studyhow climate change affects biodiversity, and how biodiversityrelates to the health of ecosystems By necessity, most of thesestudies have been theoretical (Kerr and Packer, 1998; Ives etal., 1999) and/or focused on experiments at the level of studyplots, which track diversity changes in response to environ-mental changes or treatments that take place over months,years, or at best decades (see, e.g., Brown et al., 1997; Chapin

et al., 2000; Tilman, 2000; Reich et al., 2001; Tilman et al.,2001) Difficulties arise in scaling the results from small studyplots up to the landscape, ecosystem, and biome levels (Loreau

et al., 2001) A further difficulty lies in understanding how sults obtained over short time scales compare with the natu-ral baseline of variation inherent over ecologically long timescales: hundreds to thousands to millions of years To studythis question, other researchers have focused on tracking

re-O N E

Climate Change, Biodiversity, and Ecosystem Health

The Past as a Key to the Future

A N T H O N Y D B A R N O S KY

University of California, Berkeley

ecosystem changes across major climatic transitions, such as

those at the Paleocene-Eocene boundary (Wing, 1998), in the

early Oligocene (Prothero and Heaton, 1996; Barnosky and

Carrasco, 2002), across the middle Miocene climatic optimum

(Barnosky, 2001; Barnosky and Carrasco, 2002), and across

the Pleistocene-Holocene transition (Graham and Grimm,

1990; Graham, 1992; Webb, 1992; FAUNMAP Working Group,

1996) To link across temporal scales, some studies have taken

a comparative approach, which examines how flora and fauna

responded to climate changes over varying time scales from

years to decades to centuries to thousands or millions of years

(Brown et al., 2001; Barnosky et al., 2003) A missing piece of

the puzzle, however, has been data sets that allow scientists

to track changes in biodiversity through multiple climatic

fluctuations over hundreds of thousands of years in one

geo-graphic locality

This book offers one such data set, in the form of more than

20,000 identified specimens of fossil vertebrates distributed

over more than 200,000 years, spanning the time from

ap-proximately 1,000,000 to at least 780,000 years ago The

spec-imens come from more than 26 fossil localities within

Por-cupine Cave, in the high Rocky Mountains of South Park,

Colorado (see chapter 2 for locality details) They span at least

two glacial-interglacial transitions as well as smaller-scale

climatic fluctuations within glacials and interglacials The

deposits also seem to bracket a major transition in the

period-icity of glacial-interglacial cycles, from a 41,000-year rhythm

in the early Pleistocene to a 100,000-year rhythm that was

firmly in place by 600,000 years ago Therefore it is possible to

track a single ecosystem through climate changes of variable

intensity and to assess the biodiversity response, which is one

goal of this book However, an equally important goal has

been to make the data available to future researchers in a way

that can facilitate additional analyses

Part 1 provides relevant background information on

Porcu-pine Cave, the fossil deposits themselves, and the modern

en-vironment of South Park Part 2 provides the basis for species

identifications (which are critical in assessing the quality of

the data and what it can be used for) as well as summaries

of actual numbers of specimens representing each species

(which are necessary for many ecological analyses) Part 3

fo-cuses on faunal dynamics and how the fossil information

ap-plies to understanding the effects of climatic warming on

bio-diversity The nature of the data makes it possible to examine

how climate change affected biodiversity in terms of trophic

and size structure, species richness, species composition, and

population change

An overriding impetus for this effort has been the need to

establish a baseline that will allow clear recognition of

disrup-tions to natural biodiversity caused by human-induced global

warming An initial priority is to assess how global warming

indicated by the middle Pleistocene glacial-interglacial

tran-sitions compares with rates of warming that are currently

un-der way, those that are predicted, and those that have occurred

throughout geological time

Current warming rates have long been recognized to bevery fast, and projected rates exceed rates inferred for at leastthe last 100,000 years (e.g., Schneider et al., 1992; Jackson andOverpeck, 2000) But exactly how anomalous are these fastmodern rates in comparison with the many changes in warm-ing rates that ecosystems have experienced and evolved withinover the past thousands and millions of years? Determiningthis is not as straightforward as it sounds, because rates ofchange typically are computed over differing time intervals.This has been shown to be a problem in studies of evolution-ary rates, for example, where there is an inverse relationshipbetween rates of evolutionary change and the length of timeover which the change is measured (Gingerich, 2001) Sedi-mentation rates show the opposite relationship: the thick-ness of sediments deposited over short time intervals under-

-6.00 -5.00 -4.00 -3.00 -2.00 -1.00 0.00 1.00

JC,DE

-3 -4 -5 -6H

5 5 5 5

I

2

global warming events plotted against the interval of timeover which the temperature change was measured Whitecircles show actual measurements taken from the followingsources: 1–130 years from figure 11 in Houghton et al (1990);90–900 years from figure 7.1 (bottom) in Houghton et al.(1990); 1000–10,000 years from figure 7.1 (middle) in Hough-ton et al (1990); 10,000–130,000 years from figure 6.12 inBradley (1999); 100,000–900,000 years from figure 7.1 (top)

in Houghton et al (1990); 1,000,000–2,000,000 years fromfigure 2 in Zachos et al (2001) Shaded circles mark rates forthe following observed, past, or projected global warmingepisodes: A, global warming measured from 1950 to 1990(lower dot: Houghton et al., 1990), and using a less conser-vative estimate of about 0.7°C from 1950 to 2000 (upper dot:Delworth and Knutson, 2000); B, Medieval Warm Period(Hughes and Diaz, 1994; Campbell et al., 1998; Broecker,2001); C, Pleistocene-Holocene glacial-interglacial transition(upper circle) (Schneider and Root, 1998); D, middlePleistocene glacial-interglacial transition (lower circle)(Raymo, 1997); E, Paleocene Methane Event, highest estimate(Katz et al., 1999); F, Paleocene Methane Event, lowestestimate (Katz et al., 1999); G, Middle Miocene ClimaticOptimum (Barnosky, 2001; Zachos et al., 2001; Barnosky andCarrasco, 2002); H, Late Oligocene Warming Event (Zachos

et al., 2001; Barnosky and Carrasco, 2002); I and J, lowest andhighest estimates, respectively, for global warming over thenext 100 years (Houghton et al., 2001)

estimates the total thickness that will accumulate over longer

periods (Kirchner et al., 2001) How then do rates of climate

change scale with the interval of time over which the climate

change is measured?

Figure 1.1 answers this question The data were compiled

from paleotemperature proxies provided mainly by oxygen

isotope curves (Barnosky et al., 2003) The shorter the interval

of time over which the temperature is measured, the faster the

per-hundred-year rate of change appears Plotting these data

in log-log space and highlighting the per-hundred-year

tem-perature change indicated for various past, present, and

pre-dicted warming rates place both the middle Pleistocene and

the current global warming crisis in perspective It is clear that

some of the major global warming events of the past 65

mil-lion years—the Paleocene Methane Event (Katz et al., 1999),

the late Oligocene Warming Event (Zachos et al., 2001;

Barnosky and Carrasco, 2002), the mid-Miocene Climatic

Op-timum (Barnosky, 2001; Zachos et al., 2001), middle

Pleisto-cene glacial-interglacial transitions (Raymo, 1997; Schneider

and Root, 1998), the Pleistocene-Holocene glacial-interglacial

transition (Schneider and Root, 1998), and the Medieval

Warm Period (Hughes and Diaz, 1994; Campbell et al., 1998;

Broecker, 2001)—define the high end of what is normal for

per-hundred-year rates of global warming Rates of change

measured since 1950 do not exceed the bounds of normalcy,

although, as in past global warming events, they help define

the high end of normal However, if any but the lowest

pre-dictions for the anticipated temperature rise by 2100 come to

pass, the rate of change would exceed any rates of change

known for the past 65 million years If the highest projectionsare borne out, the rate of change would be dramatic

In view of this fact, the faunal dynamics that characterizePorcupine Cave climatic transitions probably typify how eco-systems respond to climatic warming episodes that are at thehigh end of “natural” warming rates, but nevertheless do notexceed the range of rates that is normal for Earth Thus thefaunal responses to climate change that are detailed in the fol-lowing chapters are probably among the most pronouncedthat might be expected in naturally varying systems There-fore they may be useful as an ecological baseline againstwhich future changes can be measured As global warmingcontinues into the coming decades, changes in biodiversityand other faunal dynamics will undoubtedly occur—indeedare probably already occurring (Schneider and Root, 1998;Post et al., 1999; Pounds et al., 1999; Sæther et al., 2000; Bothand Visser, 2001; McCarty, 2001) Faunal responses compa-rable to those defined by the Porcupine Cave data do notnecessarily imply that the bounds of ecological health havebeen exceeded However, faunal responses that exceed thosedemonstrated by the Porcupine Cave data may well herald thedeath of the canary—a shift in the state of ecosystems that isunprecedented

Acknowledgments

Preparation of this chapter was partially supported by NSFgrant EAR-9909353 This chapter is University of CaliforniaMuseum of Paleontology contribution 1808

Porcupine Cave, arguably the richest source of information

in the world on Irvingtonian-age vertebrates, sits in the

(fig-ures 2.1, 2.2) Situated on the southwest rim of the highest

large intermountain basin in North America, known as South

Park, the cave is a three-tiered chamber comprising at least

600 m of passageways (figures 2.3–2.7) South Park itself lies

nearly in the center of Colorado (figure 2.1) and hosts diverse

biotic communities, some of which are unique in the lower

48 United States for their vegetational affinity to central Asia

Although humans have utilized various resources in South

Park for centuries, the basin remains sparsely populated,

with a wide variety of non-human-dominated landscapes still

intact

The entrance to Porcupine Cave overlooks a west-facing

slope that is near the ecotone between Festucca-Muhlenbergia

grassland, Pinus-Pseudotsuga needleleaf forest, and Picea-Abies

needleleaf forest (Küchler, 1964) Vegetation outside the

entrance consists of sparse stands of Pinus ponderosa

(Pon-derosa pine), Pinus edulis (pinyon pine), Pseudotsuga

men-ziesii (Douglas-fir), and Juniperus (juniper) interspersed with

Artemisia (sagebrush), Chrysothamnus (rabbitbrush),

soap-weed), Coryphantha (cactus), Opuntia (prickly pear), grasses,

and other small herbaceous plants (Barnosky and Rasmussen,

1988:269) The existing entrance is through a mine adit Before

emplacement of the adit (most likely in the 1870s), animals

would have had to enter the cave through various cracks and

fissures that were probably intermittently open and closed

Since 1985, when the first systematic paleontological

exca-vations took place at the site, crews from the Carnegie

Mu-seum of Natural History, the Denver MuMu-seum of Nature andScience (previously called the Denver Museum of NaturalHistory), and the University of California Museum of Paleon-tology have discovered new localities within Porcupine Cavenearly every year Chapter 4 chronicles the excavation his-tory Fifteen years of field work at the cave have yielded at least

26 different fossil localities These localities sample a widevariety of Quaternary and potentially latest Tertiary time pe-riods, and to some extent varying taphonomic situations.This chapter documents the spatial distribution of the manydifferent collecting localities and provides an overview of theirsuspected geological ages and general taphonomic settings.More details on geological age and correlation are provided inchapters 6 and 7 and in Bell and Barnosky (2000) Older pub-lications on Porcupine Cave (Barnosky and Rasmussen, 1988;Wood and Barnosky, 1994; Barnosky et al., 1996) proposed asomewhat younger age for some of the strata than is now be-lieved to be the case (see discussion of the Pit locality below).Other important Irvingtonian vertebrate paleontologicalsites from the central Rocky Mountain region include theHansen Bluff sequence in the San Luis Valley, Colorado (Rogers

et al., 1985, 1992) and the SAM Cave deposits in north-centralNew Mexico (Rogers et al., 2000) These sites are of particularinterest in yielding paleomagnetic, radiometric, palynologi-cal, and invertebrate paleontological data associated with thefossil vertebrates (Rogers et al., 1985, 1992, 2000; Rogers andWang, 2002) Specimens of vertebrate fossils and numbers ofspecies are sparse from Hansen Bluff (Rogers et al., 1985, 1992)and moderate in abundance at SAM Cave, with the latter in-cluding 2 species of amphibians, 3 of reptiles, approximately

10 of birds, and approximately 30 of mammals, distributedthrough 14 collecting localities (Rogers et al., 2000)

T W O

The Pleistocene Fossils of Porcupine Cave, Colorado

Spatial Distribution and Taphonomic Overview

Spatial Distribution of Localities

The earliest map of Porcupine Cave was published by Parris

(1973) During the course of exploration for fossils, new

pas-sageways were discovered in the late 1980s and 1990s, and

accordingly renewed mapping efforts were undertaken by

members of the Colorado Grotto and Front Range Grotto

of the National Speleological Society The updated map

pro-duced for this book (figure 2.3) was surveyed and drafted by

teams led by Hazel Barton, Kirk Branson, and Don Rasmussen,

and it shows the location of the major fossil localities

discov-ered as of 2000 that are mentioned in this book In some cases

different institutions excavated at the same locality and each

institution assigned its own locality number Table 2.1 presents

the resulting synonymies (i.e., the same locality represented

by two or more different numbers), keys the names of the

var-ious localities to figure 2.3, and summarizes the geological age

interpreted for each locality

General Taphonomic Setting

Hundreds of thousands of vertebrate fossils have been covered from Porcupine Cave, with identifiable, curated spec-imens numbering in the thousands from such single localities

re-as the Pit and Velvet Room (DMNH 644) Why were so manybones preserved?

Accumulation of Bones

At least three vectors of bone accumulation seemed to havebeen active when the cave was open in the early and middlePleistocene The most important of these was probably the

propensity of wood rats (Neotoma spp.) to collect random

items to incorporate into their middens (Betancourt et al.,1990) Collected items include carnivoran fecal pellets andraptor regurgitation pellets, which are frequently laden withthe bones of small vertebrates (especially mammals) that the

predators ate Wood rats also dragged isolated bones of large

mammals into their midden sites, including bones as large as

deer humeri or jaws, horse teeth, and podial elements of horse

and elk-sized perissodactyls and artiodactyls During

excava-tion of Porcupine Cave by CM crews in the 1985–89 field

sea-sons, bushy-tailed wood rats (Neotoma cinerea) were observed

as far back into the cave as the Pit, and fecal pellets, active

nests and middens, and urine deposits (termed “amberat” by

some authors) were observed in localities throughout the

cave Active or recently active middens within the cave

con-tained diverse plant remains (e.g., sticks, twigs, seeds), teeth

of cows, raptor pellets and carnivoran scat (probably coyote),

and in one case a dead wood rat In fact, all of these items were

contained in a single nest approximately 15 m inside the cave

Direct evidence that this activity went on for decades

in-cludes the recovery of a tobacco can with a note dated to 1939

from an active wood rat nest in the Velvet Room when the

room was first opened to humans in 1986; the can had been

dragged in by wood rats from the cave entrance (see

Neotomamidden (the Trailside Nest) about 100 m north of

the mine adit (at the site known as Trailside Entrance)

con-firms that wood rats have actively collected in the area for at

least millennia (Barnosky and Rasmussen, 1988)

Evidence that wood rat activities contributed to rich mulations of bone throughout the cave during the early andmiddle Pleistocene includes the following:

accu-1 Abundant fossils, purportedly of at least five different

species of Neotoma, including N cinerea, are present in

most of the excavated deposits in the cave

2 Within Velvet Room strata excavated by CM werefossilized middens, characterized by moderatelyindurated, tannish white layers that feature abundant

casts of the shape and size of Neotoma fecal pellets

(figure 2.8)

3 Many of the larger bones bear paired incisor gnaw

marks of a size appropriate to Neotoma.

4 The vast majority of the fossil bones are of a size that isappropriate for wood rats to incorporate into theirmiddens

5 Most of the fossils are of small mammals, withoverrepresentation of teeth, skulls with broken crania,mandibles, and other elements resistant to digestion.These characteristics, coupled with etching by stomach acids

in some cases, imply that some bones passed through the

di-Porcupine Cave Entrance

the distance

gestive tracts of carnivorans and raptors and were contained

in fecal or regurgitation pellets before being dragged into

the cave In situations in which rocky outcrops provide roosts

for raptors and/or denning areas for mammalian carnivores

within the foraging range of wood rats (as at Porcupine Cave),

wood rat middens include many bone-laden pellets Over time,

much of the organic matter except bones decays, and the

resulting deposits can be exceptionally rich in fossils (Hadly,

1999)

The second most important collection vector may have

been the direct activity of mammalian carnivores either

tak-ing prey into the cave or dytak-ing there This mode of collection

applies especially to some of the few bones that are too large

to have been dragged by wood rats Fossils of bears, badgers

and other mustelids, coyotes, and wolves have been found

in Porcupine Cave Pleistocene denning activity is suggested

by the presence of dentitions of juvenile coyotes Extant

rela-tives of all these carnivores use caves as places to bring

car-casses of small mammals or parts of large animals that they

subsequently gnaw or eat From 1985 to 1991 it was not

un-common to hear coyotes howling near the cave; signs of black

bear activity (e.g., tracks, overturned rocks) were infrequently

evident near the cave entrance; and 20 m inside the cave the

nearly complete carcass of a recently killed and partially eaten

rabbit was found in 1986 Thus extant mammalian carnivores

clearly use the cave, and there is no reason to suspect that their

extinct relatives did not also use it when adequate entrances

were available

Very rarely in Porcupine Cave are fossil animals much larger

than rodents represented by bones of a substantial portion

of the skeleton An exception is a single cranium of the camel

Camelops,which was recovered by DMNH crews from Tobacco

Road (figure 2.3) Because of its size, the skull possibly

repre-sents an animal that either fell into the cave through an

in-termittently open sinkhole, wandered in and could not find

its way out, or was dragged in as a partial carcass by a large

carnivore such as a bear

Preservation of Bones

Cave environments protect bones from decay because

tem-perature fluctuations are slight, temtem-peratures are relatively low

(thus inhibiting bacterial activity), and caves frequently are

formed in limestone, which keeps groundwater at pH values

conducive to bone preservation Porcupine Cave is no

excep-tion The cave appears to have had entrances large enough for

wood rats to enter during many periods between about 2 Ma

and at least 300 Ka ago Some openings sufficient for larger

animals to pass through probably also existed intermittently

During times of open entrances the bones accumulated Then

all entrances to the cave were apparently sealed between

some-time in the Irvingtonian (based on the age of the youngest

fossil bones) and the late 1800s, when miners opened an adit

that intersected the Gypsum Room Since that time, bones

have once again begun to accumulate in the cave from the

processes described previously, but these are easy to

differen-tiate from the fossil bones because they are on the surface ofthe cave floor and look much fresher

Implications for Ecological Interpretations

The collection vectors described earlier mean that the sample

of Pleistocene bones represents animals that lived mostlywithin a 5- to 18-km radius of Porcupine Cave Wood ratsgenerally collect within 50 m of their nest; raptors usually col-lect their prey within about 5 km of the sites where they re-gurgitate most of their pellets; and mammalian carnivoressuch as coyotes and badgers typically hunt within 5 km oftheir dens (Hadly, 1999) Porder et al (2003) found that inYellowstone Park, the bones from two deposits (Lamar Caveand Waterfall Locality) that are taphonomically similar toPorcupine Cave came from within an 8- to 18-km radius of thefossil accumulations

The derivation of most of the fossil bones from raptor lets and mammalian carnivore scats means that the samplerepresents primarily what the predators hunted Typically thediets of predators such as coyotes, hawks, and owls reflectthose small mammals and birds that are abundant on thelandscape; that is, they eat what is out there, rather than se-lectively looking for a certain species This situation results

pel-in a correlation between rank order abundance of smallmammal species identified in the pellets and scats and rankorder abundance of species in the living community, espe-cially if the predators included a range of both diurnal andnocturnal hunters (Hadly, 1999) The range of mammalianpredators that ultimately collected most of the PorcupineCave specimens potentially included fishers, weasels, ermines,black-footed ferrets, minks, wolverines, badgers, skunks,coyotes, wolves, foxes, bears, bobcats, and cheetahs Raptorsand other avian predators or scavengers potentially includedgolden eagles, hawks, ravens, falcons, kestrels, great hornedowls, and snowy owls Fidelity between fossil assemblages andthe communities they sample has been demonstrated in situ-ations taphonomically similar to Porcupine Cave (Hadly, 1999;Porder et al., 2003) Observations of the modern fauna aroundthe cave confirm that there is gross correspondence in rankorder abundance of kinds of species that characterize the re-gion today and those represented in the fossil deposits Forexample, the most commonly sighted small mammals are

Spermophilusspp (ground squirrels), and individuals of thatgenus are most common as fossils Voles likewise occur inhigh abundance in the modern environment and in the fos-sil deposits

Time averaging, or the degree to which a given localitylumps together animals that lived at widely different times(up to thousands of years, for example), is notoriously diffi-cult to assess in cave deposits (Graham, 1993; Gillieson, 1996)

In late Holocene deposits that are somewhat analogous tothose of the Pit locality in Porcupine Cave, stratigraphic levelsaveraging 10–30 cm were found to represent time spans fromabout 200 to 1000 years (Hadly, 1999; Hadly and Maurer,2001) This degree of time averaging is probably a best-case

F I G U R E 2 3 Map of Porcupine Cave Left side = west half; right side =east half Left side and right side slightly overlap (Cartography[including figures 2.4–2.7] by Hazel Barton, from a Silva/Sunto andtape survey done by Evan Anderson, Hazel Barton, Michael Barton,Beth Branson, Kirk Branson, Greg Glazner, Mike Grazi, Ted Lappin,Fred Luiszer, Emma Rainforth, Don Rasmussen, Vi Shweiker, and KenTiner Collecting sites labeled by A D Barnosky and C J Bell.)

scenario for Porcupine Cave The worst-case scenarios are

sit-uations like the Generator Dome locality in the vicinity of a

back-dirt pile, in which middle Pleistocene fossils were

re-covered tens of centimeters below the surface alongside

mod-ern debris (e.g., a match), indicating mixing of strata by either

animals or humans

These clearly different amounts of time averaging in

differ-ent localities, plus some differences in the degree to which

the three collection vectors noted previously produced the

bones in a given locality, preclude generalized interpretations

of the “Porcupine Cave fauna.” Instead, the approach taken in

this book is to specify from which localities fossils came in the

systematic descriptions of the included taxa, and to suggest

ecological interpretations only for those localities for which

we have adequate sampling, temporal control, and

appropri-ate taphonomic history We emphasize that subsequent

treat-ments of these fossils must take into account the spatial andtemporal provenance of the different localities In fact, Porcu-pine Cave is not “a locality” in the Rocky Mountains; it is acave that contains multiple, closely spaced, but temporallydistinct localities

Description of Localities

Most of the material described in this book was recoveredfrom seven localities: the Badger Room, Fissure Fill A, the Gyp-sum Room, the Pit, and three spatially distinct excavations inthe Velvet Room (Mark’s Sink, DMNH 644, CM 1927 / UCMPV93175) Of these, the most closely studied have been the Pitand DMNH 644, because those localities contained stratifiedsequences composed of multiple stacked layers and yieldedthousands of specimens The Pit locality provides the mainbasis for interpreting the effects of middle Pleistocene envi-ronmental change on ecology and evolution Material fromthe Badger Room has also been well studied and syntheticallyinterpreted Information from DMNH 644 is included inas-much as is possible, but as of this writing the locality is stillundergoing analysis, and complete results are not expected to

be available for several more years Fossils from other localitiesgenerally were analyzed only when a contributing author had

a particular interest in them Specimens from incompletelystudied localities are reported in the systematic treatments,but they contribute less to the ecological and evolutionary in-terpretations that form the last third of this book The sheervolume of material makes it impractical to provide a detailedstudy of all localities in this book This fact, and the new dis-coveries that come to light each field season, inevitably meanthat much new information remains to be reported by futuregenerations of investigators

Relevant information about each locality is presented in thefollowing sections, with localities arranged alphabetically Un-less otherwise noted, specifics of the taphonomic situationsare unknown Most UCMP samples represent subsamples ofmaterial that was collected by CM crews

Badger Room (Figure 2.3, Site 1)

but no data were recorded about depth below surface, giventhe nature of the deposits, which made such informationmeaningless in terms of time relationships Because screen

in figure 2.3: A–A′, D–D′, and F–F′

ta b l e 2.1Names and Locality Numbers of Major Vertebrate Fossil Sites of Porcupine Cave

Designation in Synonymous Locality Figure 2.3 Institutional Numbers Comments and Approximate Age

UCMP V93176

DMNH 1346

DMNH 1345UCMP V94014

UCMP V93178

DMNH 1342UCMP V93179

as Irvingtonian

DMNH 1344UCMP V98022

component

levels alongside a modern match

KU CO-121UCMP V93174

SE Corner

UCMP V93177

thought to date near 800 Ka, but are potentially as young as

250 Ka Levels 4–12 probably date to somewhere between

780 and 950 Ka Level 14 is probably younger than 1 Ma

Locality)

100 m north

of adit entrance

washing was not systematically employed, small specimens

are probably underrepresented in the CM and UCMP

collec-tions Material housed at the DMNH was collected by DMNH

crews using similar techniques, plus screen washing, during

field seasons between 1987 and 2000 DMNH extracted less

comm., 2001)

TA P H O N O M Y

Mammalian carnivores may have used this locality as a den

site and dragged in some of the bones, as indicated by gnaw

marks on some specimens, fairly abundant representation of

badger and coyote specimens (including at least one juvenile

coyote), and the abundance of rabbit bones (Anderson, 1996)

However, many of the rabbit limb bones are unbroken and

show no evidence of gnawing, which led Shabel et al

(chap-ter 22) to suggest that undevoured carcasses were left in the

cave by carnivores Abundant jaws of rodents (ranging in size

from marmots to voles) and other similarly sized specimens

indicate that at least some of the fossils derive from carnivore

fecal material or raptor pellets, many of which may have been

dragged in by wood rats

Badger Room Dome (Figure 2.3, Site 2)

LO C A L I T Y N U M B E R

DMNH 1351

C O L L E CT I O N P R OTO C O LS

This material represents a surface collection intermixed with

an ancient roof fall that collapsed and sealed a previous trance to the cave Some of the large rocks and sediment fellinto the Badger Room

en-Come-A-Long Room (Figure 2.3, Site 3)

older than Pit levels 1–4 and may be in part younger thanany Pit level Detailed studies are needed to confirm this

appear younger than 780 Ka Paleomagnetic data suggest areversal in level D Therefore, levels D–M may date tobetween 780 Ka and 1 Ma

Chance Pit

Will’s Hole

UCMP V97002note: Supporting evidence for geological ages is presented by the references cited in the text.